This proposal is Project 4 of a P50 application for the proposed Center for Interdisciplinary Research in Nicotine Addiction (CIRNA). Development of robust new medications for smoking cessation is hampered in part by inefficiency in how they are initially screened for efficacy in humans. Building on our prior work, this proposal is aimed at improving the sensitivity of procedures for screening medication efficacy. Within practical limits, such screening should simulate clinical trial methods, including use of smoking abstinence as the primary index of medication response, which may require study participants who are motivated to abstain, at least briefly. In prior research, we found that medication (NRT patch) increased abstinence over one week of use in treatment seekers, those with high intrinsic quit motivation, but extrinsic quit motivation via reinforcement for abstinence made no difference. Although informative, our findings are preliminary until they can be cross-validated with other cessation medications, which is the focus of the current project. Study 1 will examine in 200 smokers whether these results generalize to another effective medication, varenicline, by using the same mixed, cross-over design, with treatment seeking status and abstinence reinforcement as 2x2 between-subjects factors, and varenicline vs placebo as a within-subject factor. We predict that varenicline (vs placebo) will increase abstinence (i.e., show efficacy) over one week in treatment seekers more than in non-treatment seekers, confirming the importance of intrinsic quit motivation for screening. Study 2 will determine whether use of treatment seekers in such screening is both sensitive and specific. Sensitivity will be indicated if our procedure shows efficacy with another medication known to aid cessation in clinical trials, bupropion (i.e., a positive control), while specificity will be indicated if our procedure shows no efficacy in a medication known not to be effective for cessation in clinical trials, in this case modafinil (i.e., a negative control). Treatment seekers (n=100) will receive each medication and placebo in a within-subject cross-over design. We predict that, compared with placebo, bupropion will increase abstinence. Efficacy of modafinil, which is not expected, would question our procedure's specificity. Secondarily, both studies will examine individual differences and medication effects on symptoms of early abstinence (craving, withdrawal, negative and positive affect, smoking reward) to determine possible mechanisms of efficacy. Results will provide immediate directions for improving the screening of novel cessation medications, the project's longrange objective and a goal with very high public health significance.